Reply to post: If it can work ... When?

Google: We don't have a quantum computer yet, but we have a compiler

Milton Silver badge

If it can work ... When?

Notwithstanding some crayoned drivel by 'Rebel Science' (FYI "I personally cannot understand something" is NOT the same as "Science has got it wrong") I'm still wondering when quantum computing will exit the fusion energy scenario - always 10 years away - and actually achieve something of value better and faster than traditional silicon.

Working on a molecular modelling system some years ago we discovered that a bunch of GPUs[1] orchestrated in parallel by a smaller number of fast CPU cores became surprisingly effective if you broke down the tasks the right way - the sweet spot for us was to realise that "good enough" simulations could be run effectively orders of magnitude faster than "perfectly correct" ones.

Perhaps the tolerance of approximation makes this a particularly good candidate for quantum computing ... But it'll be interesting to see when qubits can produce results as good as what we were getting from a relatively cheap (<six digits £) pile of silicon, and at what price.

[1] Uncontroversial now, of course, but startling to us back then. I personally learned a lot about the significance of correctly defining the "atomic task" for GPU code, that is, the smallest individual task you'd set, after which further tasks became interlocking, summarised, aggregated etc.

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon


Biting the hand that feeds IT © 1998–2019